Animal behavior recording and collecting device

Abstract

The invention discloses an animal behavior record collecting device which comprises a three-dimensional supporting frame, a transparent experiment platform is arranged on the three-dimensional supporting frame, a circle of baffles are arranged on the transparent experiment platform, and at least one experiment target is placed in an area defined by the baffles. The dual-camera module comprises a first camera and a second camera, and the first camera and the second camera respectively collect behavior data of the experimental target from the upper part and the lower part of the baffle plate; and the video acquisition and processing unit is electrically connected with the first camera and the second camera respectively. According to the device, through high-precision three-dimensional behavior analysis, the limitation of a traditional device is broken through, three-dimensional behaviors which cannot be captured by the traditional device are quantified, and through a high-frame-rate look-up camera and a wide-angle look-up camera, a macroscopic track is captured from the look-up angle, and microscopic details are recorded from the look-up angle.

Description

Technical Field

[0001] This invention relates to the field of animal behavior technology, specifically to an animal behavior recording and acquisition device. Background Technology

[0002] Currently, mainstream animal behavior recording devices are mainly used to quantify the movement, exploration, and social behaviors of laboratory animals (such as mice). These include traditional single-camera devices for open field tests and elevated plus mazes. The open field test uses a square or circular open area (usually 40×40 cm), with a single top-view camera recording the mouse's free movement, collecting and analyzing the movement trajectory, including total distance traveled and time spent in the central area. The elevated plus maze consists of a cross-shaped elevated platform (two open arms + two closed arms), with a top-view camera recording the mouse's choices between the arms, collecting and analyzing the mouse's activity time and distance in different arms, and the balance between exploration motivation and risk avoidance. These behavioral recording devices rely on software to extract the mouse's outline using background subtraction algorithms and calculate two-dimensional coordinates. Due to the use of a single camera, there is a single-view blind spot, which can lead to the omission of key behaviors (such as foot movements and abdominal posture). Furthermore, these devices cannot distinguish three-dimensional postures, making it difficult to quantify fine movements (such as whisker twitching and limb asymmetry), and in multi-mouse experiments, occlusion can easily lead to individual tracking errors. Summary of the Invention

[0003] To address the aforementioned problems in existing technologies, this invention provides an animal behavior recording and acquisition device. It aims to overcome the limitations of traditional devices through high-precision three-dimensional behavior analysis, quantifying three-dimensional behaviors that traditional devices cannot capture. Utilizing a high-frame-rate upward-looking camera and a wide-angle downward-looking camera, it captures macroscopic trajectories from a top-down perspective and records microscopic details from an upward-looking perspective. This invention can quantify the difference between active social interaction and passive contact, reveal the mechanism of social motivation deficiency in ASD mice, and continuously record dynamic changes in movement patterns and social abilities. Through machine learning, it identifies abnormal behaviors, enabling early detection of disease phenotypes, and provides a "high-precision, high-throughput" analytical tool for the study of the mechanisms of neuropsychiatric diseases and drug development.

[0004] The technical solution provided by this invention is as follows: An animal behavior recording and data collection device, which is special in that: The device includes a three-dimensional support frame, on which a transparent experimental platform is mounted. A ring of baffles is mounted on the transparent experimental platform, and at least one experimental target is placed within the area enclosed by the baffles. The dual-camera module includes a first camera and a second camera, which respectively collect behavioral data of the experimental target from above and below the baffle. A video acquisition and processing unit is electrically connected to the first camera and the second camera respectively. The video acquisition and processing unit is used to receive behavioral data acquired by the first camera and the second camera and perform data processing.

[0005] Furthermore, a calibration plate is laid at the bottom of the baffle, and the calibration plate is a 9×7 black and white checkerboard pattern printed on a transparent PET film.

[0006] Furthermore, the first camera is a Sony IMX585 model, and the second camera is a Basler ace 2 Pro model.

[0007] Furthermore, the video acquisition and processing unit includes a video acquisition box, which supports synchronous data acquisition by the first camera and the second camera.

[0008] Furthermore, the three-dimensional support frame is made of aluminum alloy or carbon fiber and includes vertical columns and horizontal beams, while the transparent experimental platform is made of acrylic sheet. The three-dimensional support frame is equipped with a top camera bracket and a bottom camera bracket, which are used to install the first camera and the second camera, respectively.

[0009] Furthermore, the experimental target was mice.

[0010] Compared with the prior art, the beneficial effects of the present invention are: 1. From "rough statistics" to "refined analysis": Traditional single cameras only provide macroscopic behavioral statistics, while this device achieves a leap from "trajectory description" to "neural mechanism inference" through "multi-dimensional data fusion" (three-dimensional coordinates + micro-motion spectrum).

[0011] 2. From "Manual Dependence" to "Automated Intelligence": Deep learning-driven unsupervised behavior classification (such as t-SNE clustering) can automatically discover new behavioral patterns that cannot be defined by traditional metrics (such as ASD-specific social hesitation states).

[0012] 3. From "single scenario" to "ecological simulation": It supports complex environments (such as social mazes and multi-mouse coexistence), provides more natural behavioral data, and improves the reliability of basic research to clinical application.

[0013] 4. High-precision capture of behavioral indicators in laboratory mice Social deficits: Three-dimensional social distance dynamics: quantifying the decrease in interaction between experimental mice and their peers (e.g., an increase in average social distance of 20%).

[0014] Body orientation analysis: ASD mice more frequently turn their backs to their companions (the proportion of mice with a body angle >120° is increased).

[0015] Repetitive stereotyped behaviors: Foot movement periodicity: Detects excessive grooming or stereotyped stepping (frequency >10Hz) by looking up.

[0016] Three-dimensional trajectory closure: the repeatability index of movement paths in ASD mice (e.g., Lempel-Ziv complexity decreased by 30%).

[0017] 5. Unique advantages of multi-mouse interaction scenarios Individual differentiation and tracking: Even when multiple mice overlap, individuals can be accurately distinguished by foot features (such as footprint size and pressure distribution) from an upward viewing angle.

[0018] Social hierarchy analysis: Combining three-dimensional spatial location and movement direction, we can identify social avoidance patterns in ASD mice (such as actively moving away from the center of the group). Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the animal behavior recording and acquisition device in an embodiment of the present invention.

[0020] The attached figures are labeled as follows: 1-Three-dimensional support frame, 2-Baffle, 3-First camera, 4-Second camera, 5-Video acquisition and processing unit. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this application, not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0022] Therefore, the detailed description of the embodiments of this application provided below with reference to the accompanying drawings is intended merely to illustrate selected embodiments of this application and is not intended to limit the scope of protection claimed by this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0023] It should be understood that in the description of embodiments of the present invention, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, features defined with "first," "second," etc., may explicitly or implicitly include one or more of the stated features.

[0024] In the description of the embodiments of the present invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection, an electrical connection, or a connection that allows for mutual communication; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in the embodiments of the present invention according to the specific circumstances.

[0025] See Figure 1 The present invention provides an animal behavior recording and collection device, including a three-dimensional support frame 1, on which a transparent experimental platform is provided, and a ring of baffles 2 is provided on the transparent experimental platform. At least one experimental target is placed in the area enclosed by the baffles 2. In this embodiment, the experimental target is a mouse.

[0026] The dual-camera module includes a first camera 3 and a second camera 4, which respectively collect behavioral data of the experimental target from above and below the baffle 2.

[0027] The video acquisition and processing unit 5 is electrically connected to the first camera 3 and the second camera 4 respectively. The video acquisition and processing unit 5 is used to receive the behavioral data acquired by the first camera 3 and the second camera 4 and perform data processing.

[0028] Optionally, a calibration plate is laid at the bottom of the baffle 2. The calibration plate is a 9×7 black and white checkerboard pattern (square size 2×2cm) printed on a transparent PET film.

[0029] It should be noted that the calibration algorithm in this embodiment is an improved Zhang Zhengyou calibration method based on OpenCV, which supports nonlinear distortion correction (radial distortion coefficients k1, k2, tangential distortion p1, p2). Dual cameras simultaneously capture multi-angle images of the chessboard pattern (at least 20 sets). The camera intrinsic parameters (focal length \(f_x, f_y\), principal point \(c_x, c_y\)) and extrinsic parameters (rotation matrix \(R\), translation vector \(T\)) are calculated. A perspective transformation matrix is ​​generated to map the dual-viewpoint pixel coordinates to a unified world coordinate system (XYZ).

[0030] Optionally, the first camera 3 is a Sony IMX585 sensor, and the second camera 4 is a Basler ace 2 Pro sensor. The parameters of the first camera 3 are: 4K resolution (3840×2160), 60fps, wide-angle lens (FOV 120°). The parameters of the second camera 4 are: Basler ace 2 Pro, global shutter, 1080p@120fps, with a macro lens. It captures macroscopic trajectories from a top-down perspective and records microscopic details (such as the contact area of ​​a foot) from a bottom-up perspective.

[0031] Optionally, the video acquisition and processing unit 5 includes a video acquisition box, which supports synchronous data acquisition between the first camera 3 and the second camera 4. The video acquisition box features a PCIe interface, dual CameraLink inputs, and supports simultaneous 4K@60fps + 1080p@120fps acquisition. It has a built-in NVMe SSD array (maximum capacity 16TB) and supports H.265 lossless compression (compression ratio 10:1). This ensures strict synchronization of the two video streams, avoiding 3D reconstruction errors caused by time misalignment.

[0032] The video acquisition and processing unit 5 can also perform data processing and analysis, as follows: Step 1: Video capture box processes the video. 1. Synchronous acquisition: Through hardware signals, the first camera 3 and the second camera 4 are forced to expose at the same microsecond, ensuring that the time of each frame is strictly aligned.

[0033] 2. Lossless Compression: The video is compressed using H.265 lossless compression, saving space without losing any image details and ensuring that every pixel value is accurate.

[0034] 3. High-speed storage: The compressed dual-channel video streams are written to the built-in NVMe SSD array at high speed to ensure no frame loss during long-term, high-frame-rate recording.

[0035] Output: Two perfectly synchronized, lossless high-definition video files are obtained.

[0036] Step 2: Perform dual-view checkerboard calibration 1. Obtain synchronized calibration video: Place a checkerboard calibration board in the mouse activity area, and use the above system to record a video of the checkerboard being seen simultaneously by the two cameras from two different shooting angles.

[0037] 2. Extract synchronized image pairs: Using the hardware timestamp recorded during acquisition, accurately extract checkerboard image pairs captured at the same time from dual-channel video.

[0038] 3. Calculate camera parameters: Input multiple sets of synchronized image pairs into calibration software (such as OpenCV) to automatically calculate: Internal parameters for each camera (to eliminate lens distortion).

[0039] The relative position and angle between the two cameras (establishing a three-dimensional coordinate system).

[0040] Ultimate goal: To obtain a high-precision "dual-camera 3D measurement model" to lay the foundation for subsequent 3D reconstruction.

[0041] Step 3: Perform machine learning analysis on mouse behavior. 1. Record experimental data: Remove the chessboard grid, allow the experimental mice to move freely, and record synchronized dual-view experimental videos using a calibrated system.

[0042] 2. 3D Reconstruction: Using the "camera model" obtained in step 2, the coordinates of the mouse's skeletal joints in 3D space are reconstructed from each frame of the experimental video.

[0043] 3. Behavioral feature extraction: Input continuous three-dimensional key point data (such as the 3D trajectory of the nose tip, limbs, and tail base) into the machine learning model.

[0044] 4. Analysis and Identification: The model automatically learns and identifies specific behaviors of mice (such as time spent in the central area and social interactions), and performs quantitative analysis on them (such as frequency, duration, and intensity).

[0045] Final outcome: To achieve automated, high-precision, and quantifiable three-dimensional analysis of the behavior of laboratory mice.

[0046] Optionally, the three-dimensional support frame 1 is made of aluminum alloy or carbon fiber and includes vertical columns and horizontal beams. The transparent experimental platform is made of acrylic sheet with a frosted surface to reduce reflection.

[0047] The three-dimensional support frame 1 is provided with a top camera bracket and a bottom camera bracket, which are used to install the first camera 3 and the second camera 4, respectively.

[0048] Specifically, the main frame of the 3D support frame 1 is made of aluminum alloy or carbon fiber, which is lightweight and highly rigid. It includes vertical columns and horizontal beams, and its height is adjustable (range 50-150cm). The transparent experimental platform is made of acrylic sheet (8-10mm thick), with a frosted surface to reduce reflection, and its dimensions are 40×40cm. The top and bottom camera brackets hold dual cameras for both overhead and under-eye views, with a 65cm distance between them and the transparent platform. A three-point gravity balance design prevents image blurring caused by camera shake.

[0049] Using the animal behavior recording and collection device provided by this invention includes the following steps: (1) Turn on the first camera 3 and the second camera 4, adjust the camera angle, and adjust parameters such as clarity and focus. The first camera 3 and the second camera 4 are connected to the video capture box via HDMI cable and connected to the video recorder.

[0050] (2) Lay out the calibration plate, turn on the recording button, and the upper and lower cameras start recording synchronously to complete the parameter calculation and coordinate registration of the dual cameras.

[0051] (3) After recording the checkerboard pattern of the calibration board for 20-30 seconds, remove it and place the circular baffle on the recording plane to ensure that the space inside the baffle can be recorded by the camera.

[0052] (4) Place the mice in the experiment according to the experimental requirements and record the results.

[0053] The animal behavior recording and acquisition device provided by this invention aims to overcome the limitations of traditional methods through high-precision three-dimensional behavior analysis, quantifying three-dimensional behaviors (such as standing height and body curling curvature) that traditional methods cannot capture. It records transient micro-movements (such as foot tremors, whisker twitches, and blinking reactions) using a high frame rate (120fps) upward-looking camera, with a sensitivity reaching sub-second levels (<10ms). (For example, in autistic mice with ASD, it can quantify the frequency, amplitude, and timing characteristics of stereotyped behaviors such as repetitive grooming, accurately linking genotype and phenotype). It can also quantify the difference between active social interaction (such as nose touch and following) and passive contact (such as collision), revealing the mechanism of social motivation deficiency in ASD mice.

[0054] This device is not only a behavioral recording tool, but also a core infrastructure for understanding brain function, decoding disease mechanisms, and advancing precision medicine. This invention has the following technical effects: 1. From "rough statistics" to "refined analysis": Traditional single cameras only provide macroscopic behavioral statistics, while this device achieves a leap from "trajectory description" to "neural mechanism inference" through "multi-dimensional data fusion" (three-dimensional coordinates + micro-motion spectrum).

[0055] 2. From "Manual Dependence" to "Automated Intelligence": Deep learning-driven unsupervised behavior classification (such as t-SNE clustering) can automatically discover new behavioral patterns that cannot be defined by traditional metrics (such as ASD-specific social hesitation states).

[0056] 3. From "single scenario" to "ecological simulation": It supports complex environments (such as social mazes and multi-mouse coexistence), provides more natural behavioral data, and improves the reliability of basic research to clinical application.

[0057] 4. High-precision capture of behavioral indicators in laboratory mice Social deficits: Three-dimensional social distance dynamics: quantifying the decrease in interaction between experimental mice and their peers (e.g., an increase in average social distance of 20%).

[0058] Body orientation analysis: ASD mice more frequently turn their backs to their companions (the proportion of mice with a body angle >120° is increased).

[0059] Repetitive stereotyped behaviors: Foot movement periodicity: Detects excessive grooming or stereotyped stepping (frequency >10Hz) by looking up.

[0060] Three-dimensional trajectory closure: the repeatability index of movement paths in ASD mice (e.g., Lempel-Ziv complexity decreased by 30%).

[0061] 5. Unique advantages of multi-mouse interaction scenarios Individual differentiation and tracking: Even when multiple mice overlap, individuals can be accurately distinguished by foot features (such as footprint size and pressure distribution) from an upward viewing angle.

[0062] Social hierarchy analysis: Combining three-dimensional spatial location and movement direction, we can identify social avoidance patterns in ASD mice (such as actively moving away from the center of the group). Compared with traditional single-camera devices, the device provided by the present invention has the advantages shown in Table 1.

[0063] Table 1 Traditional single camera This dual-camera device Advantages Two-dimensional trajectory (XY coordinates) 3D pose reconstruction (XYZ coordinates + joint angles) Quantify anxiety-related three-dimensional indicators such as standing height and torso curl curvature. Relying on manual observation of microscopic behavior Automatically detects micro-movements (such as the frequency of foot tremors and the amplitude of beard swaying when viewed from an upward angle). Precisely identify sub-second stereotyped behaviors in ASD mice (such as repetitive grooming at 0.5 Hz). Pre-labeling is required for multi-mouse experiments. Distinguishing unmarked individuals (by observing differences in foot morphology or three-dimensional movement patterns from above) Accurately track multi-mouse interactions in natural social scenarios (such as active / passive determination of attack and sniffing). Low temporal resolution (30fps) High-speed record Capture transient behaviors (such as delayed startle reflex in mice). Behavioral indicators Traditional single-view system This device ASD Research Value Social interaction frequency It can only count the time two mice are in the same frame, and cannot distinguish between active contact. Determining true social intentions by combining the three-dimensional distance of the nose tip with body orientation. More accurate identification of social motivation deficiency in ASD mice Stereotype detection Relying on manual observation or rough judgment based on trajectory closure Quantitative foot motion spectrum + body curvature angle periodicity Sub-second micro-stereotyped behaviors were detected (such as abnormal beard twitching). Developmental trajectory analysis Manual scoring of discrete time point behavior changes Continuous 3D motion pattern machine learning modeling Predicting ASD phenotypic progression and the window of opportunity for drug intervention This device, through its triple resolution enhancement of "space-time-behavior", not only overcomes the inherent shortcomings of traditional single-camera devices, but also provides a "high-precision, high-throughput" analytical tool for the study of the mechanisms of neuropsychiatric disorders (such as ASD and anxiety disorders) and drug development.

[0064] The above description is merely the preferred embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An animal behavior recording and acquisition device, characterized in that: The device includes a three-dimensional support frame, on which a transparent experimental platform is mounted. A ring of baffles is mounted on the transparent experimental platform, and at least one experimental target is placed within the area enclosed by the baffles. The dual-camera module includes a first camera and a second camera, which respectively collect behavioral data of the experimental target from above and below the baffle. A video acquisition and processing unit is electrically connected to the first camera and the second camera respectively. The video acquisition and processing unit is used to receive behavioral data acquired by the first camera and the second camera and perform data processing.

2. The animal behavior recording and acquisition device according to claim 1, characterized in that: A calibration plate is laid at the bottom of the baffle, and the calibration plate is a 9×7 black and white checkerboard pattern printed on a transparent PET film.

3. The animal behavior recording and acquisition device according to claim 1, characterized in that: The first camera is a Sony IMX585 model, and the second camera is a Basler ace 2 Pro model.

4. The animal behavior recording and acquisition device according to claim 1, characterized in that: The video acquisition and processing unit includes a video acquisition box, which supports synchronous data acquisition by the first camera and the second camera.

5. The animal behavior recording and acquisition device according to any one of claims 1-4, characterized in that: The three-dimensional support frame is made of aluminum alloy or carbon fiber and includes vertical columns and horizontal beams; the transparent experimental platform is made of acrylic sheet. The three-dimensional support frame is equipped with a top camera bracket and a bottom camera bracket, which are used to install the first camera and the second camera, respectively.

6. The animal behavior recording and acquisition device according to claim 5, characterized in that: The experimental target was mice.

Images